A psycholinguistic view on definites · indefinite definite possessive INDIVIDUAL– IC pope earth...

A psycholinguistic viewon definites

Düsseldorf, 01‐02 June, 2012

Dorothea BrennerPeter Indefrey

Heinrich Heine University Düsseldorf, Collaborative Research Centre 991 

The Structure of Representations in Language, Cognition and ScienceProject C3: Psycholinguistic Evidence for Concept Types

Semantic and typological perspectives on definites

Outline

1. Theoretical Background1.1 Assumptions1.2 Research Questions & Hypotheses

2.Psycholinguistic Experiment2.1 Paradigm2.2 Method2.3 Results

3.Summary & further research objectives

2

1. Theoretical Background: CTD (modif. version of Löbner 2011:307)

congruent determination incongruent determination 3

A Psycholinguistic View on Definites

[–U] conceptually unique [+U]

[–R]

SORTAL – SCapple stone moment humanindefinitedefinitepossessive

INDIVIDUAL – ICpope earth weather Policeindefinitedefinitepossessive

conceptuallyrelational [+R]

RELATIONAL – RCcolleague arm page ideaindefinitedefinitepossessive

FUNCTIONAL – FCmother body age birthindefinitedefinitepossessive

[–U] conceptually unique [+U]

[–R]

SORTAL – SCapple stone moment humanindefinitedefinitepossessive

INDIVIDUAL – ICpope earth weather Policeindefinitedefinitepossessive

conceptuallyrelational [+R]

RELATIONAL – RCcolleague arm page ideaindefinitedefinitepossessive

FUNCTIONAL – FCmother body age birthindefinitedefinitepossessive

1.1 Assumptions: 1) Underlying CT• Concept Types (CT)

– Concept type information of nouns is lexically stored– Most nouns have only one lexically stored concept type and corresponding frame specification

underlying concept type

A Psycholinguistic View on Definites

4A Psycholinguistic View on Definites

1.1 Assumptions: 2) Type shifts• CTs & Determination

– Each of the four concept types has a preferred contextual profile (c.f. Löbner 2011), i.e. it is used with specific “congruent” determination type (DT)

• CTs & Incongruent Determination– The interpretation of a noun used with an incongruent DT leads to a reanalysis process, so that its referential properties then match the ones required by the DT.

conceptual type shift (CT‐shift)

A Psycholinguistic View on Definites

5A Psycholinguistic View on Definites

1.1 Assumptions: Examplea)  Der Papst wohnt in Italien.

(The Pope lives in Italy.)

b)  Johannes Paul II. war ein freundlicher Papst. (John Paul II. was a friendly pope.)

– ‚Papst‘ (pope) is an IC [+U,+R]– In a) it is used with congruent determination– the indefinite article ‚ein‘ in b) requires a [–U]‐concept.incongruency between CT and DT

– the interpretation of b) requires a reanalysis process: the referential properties of the IC ‚Papst‘ have to be changed, to match the values required by the DT ‚ein‘incongruency coerces a CT‐shift

A Psycholinguistic View on Definites

6A Psycholinguistic View on Definites

1.2 Research Questions & Hypotheses• Empirical Research Questions:

– Do CTs and CT‐shifts have a measurable cognitive reality?– Can we find empirically measurable time differences in the processing of nouns used with congruent vs. incongruent  DT?

• Hypotheses & Prediction:– Congruent determination should facilitate the processing of the respective noun

– Incongruent determination leads to CT‐shifts, which should  be time‐consuming and thus slow down responses.

A Psycholinguistic View on Definites

7A Psycholinguistic View on Definites

2.1 Experiment: Paradigm• On‐line reaction time experiment with German NPs containing a combination of determiner+noun

• Lexical decision task: – Task: „Is the presented stimulus a word or a non‐word?“– triggers lexical and (flat) semantic processing

• Presentation mode: auditory

• Measured variable: reaction time (RT)via response pad

A Psycholinguistic View on Definites

8A Psycholinguistic View on Definites

2.2 Experiment: Stimuli & Method

• Participants: 96 German native speakers

• Stimuli: 

– target nouns: 80 German nouns – 20 nouns of each CT (matched by frequency and number of letters and phonemes)

– pseudo words: 80 non‐words satisfying the phonotacticrules of German

– each item was combined with each of the four determinertypes indefinite, definite, possessive, none

A Psycholinguistic View on Definites

9A Psycholinguistic View on Definites

Four combinations of CT and DT:  

2.2.1 Experiment: Stimuli

Concept typeDet. type SORTAL

[‐U, ‐R]INDIVIDUAL[+U, ‐R]

RELATIONAL[‐U, +R]

FUNKTIONAL[+U, +R]

indefinite ein Apfelan apple

ein Papsta pope

ein Arman arm

eine Muttera mother

definite der Apfelthe apple

der Papstthe pope

der Armthe arm

die Mutterthe mother

possessive sein Apfelhis apple

sein Papsthis pope

sein Armhis arm

seine Mutterhis mother

none xxxx Apfel xxxx Papst xxxx Arm xxxx Mutter

A Psycholinguistic View on Definites

10A Psycholinguistic View on Definites

Concept typeDet. type SORTAL

[‐U, ‐R]INDIVIDUAL[+U, ‐R]

RELATIONAL[‐U, +R]

FUNKTIONAL[+U, +R]

indefinite ein Apfelan apple

ein Papsta pope

ein Arman arm

eine Muttera mother

definite der Apfelthe apple

der Papstthe pope

der Armthe arm

die Mutterthe mother

possessive sein Apfelhis apple

sein Papsthis pope

sein Armhis arm

seine Mutterhis mother

none xxxx Apfel xxxx Papst xxxx Arm xxxx Mutter

congruent determinationincongruent determination

2.2.2 Experiment: Method

Each of the 160 trials consisted of 3 subsequent parts:

a fixation stimulus: „beep“

one of the three determiners or the neutral determiner stimulus (realized as 400ms white noise)

one of the 80 target words or one of the 80 pseudo words

11A Psycholinguistic View on Definites

• Mean reaction time for [±U]‐concepts

2.3.1 Experiment: Results [±U]

A Psycholinguistic View on Definites

12A Psycholinguistic View on Definites

Significant Effects:• statistically significant 

interaction effect between  determination & uniqueness (F(93)=8.09, p=.00) 

• post‐hoc comparisons show: – significant facilitation of

[+U]‐nouns by definite DT– significant facilitation of

[‐U]‐nouns by indefinite DT

2.3.2 Experiment: Results [±R]• Mean reaction time for [±R]‐concepts

A Psycholinguistic View on Definites

13A Psycholinguistic View on Definites

Significant Effects:• statistically significant interaction effect between determination & relationality(F(93)=6.76, p=.00)

• post‐hoc comparisons show:– significant facilitation of [+R]‐nouns by possessive DT

– inhibition of [‐R]‐nouns bypossessive DT

– unexpected significantfacilitation effect of [‐R]‐nounsby definite DT  

2.3.3 Experiment: Results Congruency• Mean reaction times for congruent, 

incongruent and neutral DT

A Psycholinguistic View on Definites

14A Psycholinguistic View on Definites

Significant Effects:• highly significant difference

between congruent, incongruent & nodetermination (F(94)=12,85; p= .00)

• Post‐hoc comparison shows: – significant difference

between congruent vs. incongruent determination

– no significant differencebetween incongruent vs.  nodetermination

results cannot be explainedby mere gender effect ofdetermination

3. Summary & Further Questions

• Results show evidence for– the cognitive reality of the distinction of the four concepttypes within the CTD 

– the interaction of determiner type and concept type

• Further research objectives:– Differences in the data for visual mode of speechperception?

– Processing stage (lexical or post‐lexical)?– Mechanisms & time course of processing conceptualinformation?

15Conceptual Shifts: Psycholinguistic EvidenceA Psycholinguistic View on Definites

ReferencesBölte, J. & Connine, C. M. (2004). Grammatical gender in spoken word recognition in 

German. Perception & Psychophysics 66, pp. 1018‐1032.

Goldinger, S. D. (1997). Auditory lexical decision. In: Grosjean, F. & Frauenfelder, U. H. (Eds.). A guide to spoken word recognition paradigms. [Language and Cognitive Processes, 11, 559 ‐ 567.]

Indefrey, P. and Cutler, A. (2004) Prelexical and lexical processing in listening. In:Gazzaniga, M. (ed.) The Cognitive Neurosciences III. Cambridge, MA: MIT Press, pp. 759‐774.

Löbner, S. (2011). Conceptual Types and Determination. Journal of Semantics 28, pp. 279‐333. 

Vigliocco, G., Vinson, D., Indefrey, P., Levelt, W. J. M., & Hellwig, F. (2004). Role of grammatical gender and semantics in German word production. Journal of Experimental Psychology: Language, Memory, and Cognition 30, pp. 483‐497.

A Psycholinguistic View on Definites

16A Psycholinguistic View on Definites